The rest of your point seems to hold, though; if the subsidy is predicated on reducing CO2 emissions then the equivalent of 25mpg still isn’t anything to brag about.
This is likely an overestimation, since it assumes that you’re exclusively burning coal. Electricity production in the US is about 68% fossil, the rest deriving from a mixture of nuclear and renewables; the fossil-fuel category also includes natural gas, which per your link generates about 55-60% the CO2 of coal per unit electricity. This varies quite a bit state to state, though, from almost exclusively fossil (West Virginia; Delaware; Utah) to almost exclusively nuclear (Vermont) or renewable (Washington; Idaho).
Based on the same figures and breaking it down by the national average of coal, natural gas, and nuclear and renewables, I’m getting a figure of 43 lb CO2 / 100 mi, or about 50 mpg equivalent. Since its subsidies came up, California burns almost no coal but gets a bit more than 60% of its energy from natural gas; its equivalent would be about 28 lb CO2.
This is likely an overestimation, since it assumes that you’re exclusively burning coal.
Yes, but that should be the right comparison to make. Consider two alternatives:
1) World generates N kwh + 38 kwh to fuel a Tesla to go 100 miles
2) World generates N kwh and puts 4 gallons of gasoline in a car to go 100 miles.
If we are interested in minimizing CO2 emissions, then in world 2 compared to world 1 we will generate 38 kWh fewer from our dirtiest plant on the grid, which is going to be a coal-fired plant.
So in world 1 we have an extra 80 lbs of CO2 emission from electric generation and 0 from gasoline.
In world 2 we have 80 lbs less of CO2 emission from electric generation and add 80 lbs from gasoline.
When adding electric usage, you need to “bill” it at the marginal costs to generate that electricity, which is true both in terms the price you charge customers for it and the CO2 emissions you attribute to it.
The US, China, and most of Europe have a lot of Coal in the mix on the grid. Until they scrub coal or stop using it, it seems very clear that the Tesla puffs out the same amount of CO2 as a 25 mpg gasoline powered car.
It’s true that most of the flexibility in our power system comes from dirty sources, and that squeezing a few extra kilowatt-hours in the short term generally means burning more coal. If we’re talking policy changes aimed at popularizing electric cars, though, then we aren’t talking a megawatt here or there; we’ve moved into the realm of adding capacity, and it’s not at all obvious that new electrical capacity is going to come from dirty sources—at least outside of somewhere like West Virginia. On those kinds of scales, I think it’s fair to assume a mix similar to what we’ve currently got, outside of special cases like Germany phasing out its nuclear program.
(There are some caveats; renewables are growing strongly in the US, but nuclear isn’t. But it works as a first approximation.)
Coal electric generation isn’t going away anytime soon. The only reason coal may look at the moment like it is declining in the US is because at the moment natural gas generation in the US is less expensive than coal. But in Europe, coal is less expensive and, remarkably, generating companies respond by turning up coal and turning down natural gas.
Doesn’t need to be going away for my argument to hold, as long as the relative proportions are favorable—and as far as I can tell, most of that GIC delta in coal is happening in the developing world, where I don’t see too many people buying Teslas. Europe and the US project new capacity disproportionately in the form of renewables; coal is going up in Europe, but less quickly.
This isn’t ideal; I’m generally long on wind and solar, but if I had my way we’d be building Gen IV nuclear reactors as fast as we could lay down concrete. But neither is it as grim as the picture you seem to be painting.
This isn’t ideal; …. But neither is it as grim as the picture you seem to be painting.
I would agree with that.. Certainly my initial picture was just wrong. Even using Coal as the standard, the Tesla is as good as a 25 mpg gasolilne car. For that size and quality of car, that is actually not bad, but it is best in class, not revolutionary.
As to subsidizing a Tesla as opposed to a 40 mpg diesel, for example, as long as we use coal for electricity, we are better off adding a 40 mpg diesel to the fleet than adding a Tesla. This is almost just me hating on subsidies, preferring that we just tax fuels proportional to their carbon content and let market forces decide how to distribute that distortion.
This is almost just me hating on subsidies, preferring that we just tax fuels proportional to their carbon content and let market forces decide how to distribute that distortion.
That probably is better baseline policy from a carbon minimization perspective, yeah; I have similar objections to the fleet mileage penalties imposed on automakers in the US, which ended up contributing among other things to a good chunk of the SUV boom in the ’90s and ’00s. Now, I can see an argument for subsidies or even direct grants if they help kickstart building EV infrastructure or enable game-changing research, but that should be narrowly targeted, not the basis of our entire approach.
Unfortunately, basic economic literacy is not exactly a hallmark of environmental policy.
When adding electric usage, you need to “bill” it at the marginal costs to generate that electricity
Yes, but marginal analysis requires identifying the correct margin. If you charge your car during the day at work, you are increasing peak load, which is often coal. If you charge your car at night, you are contributing to base load. This might not even require building new plants! This works great if you have nuclear plants. With a sufficiently smart grid, it makes erratic sources like wind much more useful.
Yes, but marginal analysis requires identifying the correct margin.
I do agree using the rate for coal is pessimistic.
On further research, I discover that Li-ion batteries are very energetically expensive to produce. Their net lifetime energy in production and then recycling is about 430 kWh per kWh of battery. Li-ion can be recharged 300-500 times. Using 430 recharges, amortizing production costs across all uses of the battery we see that we have 1 kWh of production energy used for every 1 kWh of storage the battery accomplished during its lifetime.
So now we have the more complicated accounting questions, how much carbon do we associate with constructing the battery vs how much with charging the battery? If construction and charging come from the same grid, we charge the same.
And of course to be fair, we need to figure the cost to refine a gallon of gasoline. Its pretty wacky out there but the numbers out there range from 6 kwh to 12 kwh. The higher numbers include quite a bit of natural gas directly used in the process, which using it directly is about twice as efficient as making electricity with it.
All in all, it looks to me like we have about 100% overhead on battery production energy, and say 8 kWh to make a gallon of gas for about 25% overhead on gasoline.
Lets assign 1.3 lbs of CO2 per kwh electric, which is 2009 US average adjusted 7.5% for delivery losses.
Then a gallon of gasoline gives 19 lbs from the gasoline + 10.4 lbs from making/transporting the gasoline.
A Tesla costs 1.3*38 = 39 lbs CO2 to go 100 miles from electric charge + 39 lbs CO2 from amortizing battery lifetime over CO2 cost or producing the battery.
Tesla = 78 lbs CO2 per 100 miles.
A 78 lbs of CO2 comes from 78⁄30 = 2.6 gallons of fuel.
So using US average CO2 load for kwh electricity, loading the Tesla with 100% overhead for battery production and loading gasoline with 34% overhead from refining, mining, and transport, we get a Tesla S about equivalent to a 38 mpg car in CO2 emissions.
That number is actually extremely impressive for the class of car a Tesla is.
Nissan Leaf uses 75% as much energy as Tesla to go 100 miles. So Leaf has same CO2 emissions as a 51 mpg car.
If we use coal for electricity these numbers change to Tesla --> 19 mpg and Leaf --> 26 mpg. The Tesla still looks good-ish for the class of car it is, but the Leaf is lousy at 26 mpg, competing with hybrids that get 45 mpg or so.
Your lithium-ion numbers match my understanding of batteries in general: they cost as much energy to create as their lifetime capacity. That’s why you can’t use batteries to smooth out erratic power sources like wind, or inflexible ones like nuclear.
I’m skeptical that it’s a good idea to focus on the energy used to create the battery. There’s energy used to create all the rest of the car, and certainly energy to create the gasoline-powered car that you’re using as a benchmark. Production energy is difficult to compute and I think most people do such a bad job that I think it’s better to use price as a proxy.
The rest of your point seems to hold, though; if the subsidy is predicated on reducing CO2 emissions then the equivalent of 25mpg still isn’t anything to brag about.
You are right I did my math wrong.
To make it a little clearer to people following along, 80 lbs of CO2 generate to move a Tesla 100 miles using coal generated electricity. 80 pounds of CO2 to move a 25 mpg gasoline car 100 miles.
I’ll address why the coal number is the right one in commenting on the next comment.
The Tesla S takes about 38 kW-hr to go 100 miles, which works out to around 80 lb CO2 generated. 14mpg would be 7.1 gallons of gasoline to go 100 miles, which works out to around 140lb CO2 generated. I couldn’t find any independent numbers for the S Performance, but Tesla’s site claims the same range as the regular S with the same battery pack.
The rest of your point seems to hold, though; if the subsidy is predicated on reducing CO2 emissions then the equivalent of 25mpg still isn’t anything to brag about.
This is likely an overestimation, since it assumes that you’re exclusively burning coal. Electricity production in the US is about 68% fossil, the rest deriving from a mixture of nuclear and renewables; the fossil-fuel category also includes natural gas, which per your link generates about 55-60% the CO2 of coal per unit electricity. This varies quite a bit state to state, though, from almost exclusively fossil (West Virginia; Delaware; Utah) to almost exclusively nuclear (Vermont) or renewable (Washington; Idaho).
Based on the same figures and breaking it down by the national average of coal, natural gas, and nuclear and renewables, I’m getting a figure of 43 lb CO2 / 100 mi, or about 50 mpg equivalent. Since its subsidies came up, California burns almost no coal but gets a bit more than 60% of its energy from natural gas; its equivalent would be about 28 lb CO2.
Yes, but that should be the right comparison to make. Consider two alternatives: 1) World generates N kwh + 38 kwh to fuel a Tesla to go 100 miles 2) World generates N kwh and puts 4 gallons of gasoline in a car to go 100 miles.
If we are interested in minimizing CO2 emissions, then in world 2 compared to world 1 we will generate 38 kWh fewer from our dirtiest plant on the grid, which is going to be a coal-fired plant.
So in world 1 we have an extra 80 lbs of CO2 emission from electric generation and 0 from gasoline. In world 2 we have 80 lbs less of CO2 emission from electric generation and add 80 lbs from gasoline.
When adding electric usage, you need to “bill” it at the marginal costs to generate that electricity, which is true both in terms the price you charge customers for it and the CO2 emissions you attribute to it.
The US, China, and most of Europe have a lot of Coal in the mix on the grid. Until they scrub coal or stop using it, it seems very clear that the Tesla puffs out the same amount of CO2 as a 25 mpg gasoline powered car.
It’s true that most of the flexibility in our power system comes from dirty sources, and that squeezing a few extra kilowatt-hours in the short term generally means burning more coal. If we’re talking policy changes aimed at popularizing electric cars, though, then we aren’t talking a megawatt here or there; we’ve moved into the realm of adding capacity, and it’s not at all obvious that new electrical capacity is going to come from dirty sources—at least outside of somewhere like West Virginia. On those kinds of scales, I think it’s fair to assume a mix similar to what we’ve currently got, outside of special cases like Germany phasing out its nuclear program.
(There are some caveats; renewables are growing strongly in the US, but nuclear isn’t. But it works as a first approximation.)
The global installed capacity of coal-fired power generation is expected to increase from 1,673.1 GW in 2012 to 2.057.6 GW by 2019, according to a report from Transparency Market Research. Coal-fired electrical-generation plants are being started up in Europe—and comparatively clean gas-fired generating capacity is being shut down.
Coal electric generation isn’t going away anytime soon. The only reason coal may look at the moment like it is declining in the US is because at the moment natural gas generation in the US is less expensive than coal. But in Europe, coal is less expensive and, remarkably, generating companies respond by turning up coal and turning down natural gas.
Doesn’t need to be going away for my argument to hold, as long as the relative proportions are favorable—and as far as I can tell, most of that GIC delta in coal is happening in the developing world, where I don’t see too many people buying Teslas. Europe and the US project new capacity disproportionately in the form of renewables; coal is going up in Europe, but less quickly.
This isn’t ideal; I’m generally long on wind and solar, but if I had my way we’d be building Gen IV nuclear reactors as fast as we could lay down concrete. But neither is it as grim as the picture you seem to be painting.
I would agree with that.. Certainly my initial picture was just wrong. Even using Coal as the standard, the Tesla is as good as a 25 mpg gasolilne car. For that size and quality of car, that is actually not bad, but it is best in class, not revolutionary.
As to subsidizing a Tesla as opposed to a 40 mpg diesel, for example, as long as we use coal for electricity, we are better off adding a 40 mpg diesel to the fleet than adding a Tesla. This is almost just me hating on subsidies, preferring that we just tax fuels proportional to their carbon content and let market forces decide how to distribute that distortion.
That probably is better baseline policy from a carbon minimization perspective, yeah; I have similar objections to the fleet mileage penalties imposed on automakers in the US, which ended up contributing among other things to a good chunk of the SUV boom in the ’90s and ’00s. Now, I can see an argument for subsidies or even direct grants if they help kickstart building EV infrastructure or enable game-changing research, but that should be narrowly targeted, not the basis of our entire approach.
Unfortunately, basic economic literacy is not exactly a hallmark of environmental policy.
Yes, but marginal analysis requires identifying the correct margin. If you charge your car during the day at work, you are increasing peak load, which is often coal. If you charge your car at night, you are contributing to base load. This might not even require building new plants! This works great if you have nuclear plants. With a sufficiently smart grid, it makes erratic sources like wind much more useful.
I do agree using the rate for coal is pessimistic.
On further research, I discover that Li-ion batteries are very energetically expensive to produce. Their net lifetime energy in production and then recycling is about 430 kWh per kWh of battery. Li-ion can be recharged 300-500 times. Using 430 recharges, amortizing production costs across all uses of the battery we see that we have 1 kWh of production energy used for every 1 kWh of storage the battery accomplished during its lifetime.
So now we have the more complicated accounting questions, how much carbon do we associate with constructing the battery vs how much with charging the battery? If construction and charging come from the same grid, we charge the same.
And of course to be fair, we need to figure the cost to refine a gallon of gasoline. Its pretty wacky out there but the numbers out there range from 6 kwh to 12 kwh. The higher numbers include quite a bit of natural gas directly used in the process, which using it directly is about twice as efficient as making electricity with it.
All in all, it looks to me like we have about 100% overhead on battery production energy, and say 8 kWh to make a gallon of gas for about 25% overhead on gasoline.
Lets assign 1.3 lbs of CO2 per kwh electric, which is 2009 US average adjusted 7.5% for delivery losses.
Then a gallon of gasoline gives 19 lbs from the gasoline + 10.4 lbs from making/transporting the gasoline.
A Tesla costs 1.3*38 = 39 lbs CO2 to go 100 miles from electric charge + 39 lbs CO2 from amortizing battery lifetime over CO2 cost or producing the battery.
Tesla = 78 lbs CO2 per 100 miles.
A 78 lbs of CO2 comes from 78⁄30 = 2.6 gallons of fuel.
So using US average CO2 load for kwh electricity, loading the Tesla with 100% overhead for battery production and loading gasoline with 34% overhead from refining, mining, and transport, we get a Tesla S about equivalent to a 38 mpg car in CO2 emissions.
That number is actually extremely impressive for the class of car a Tesla is.
Nissan Leaf uses 75% as much energy as Tesla to go 100 miles. So Leaf has same CO2 emissions as a 51 mpg car.
If we use coal for electricity these numbers change to Tesla --> 19 mpg and Leaf --> 26 mpg. The Tesla still looks good-ish for the class of car it is, but the Leaf is lousy at 26 mpg, competing with hybrids that get 45 mpg or so.
Your lithium-ion numbers match my understanding of batteries in general: they cost as much energy to create as their lifetime capacity. That’s why you can’t use batteries to smooth out erratic power sources like wind, or inflexible ones like nuclear.
I’m skeptical that it’s a good idea to focus on the energy used to create the battery. There’s energy used to create all the rest of the car, and certainly energy to create the gasoline-powered car that you’re using as a benchmark. Production energy is difficult to compute and I think most people do such a bad job that I think it’s better to use price as a proxy.
You are right I did my math wrong.
To make it a little clearer to people following along, 80 lbs of CO2 generate to move a Tesla 100 miles using coal generated electricity. 80 pounds of CO2 to move a 25 mpg gasoline car 100 miles.
I’ll address why the coal number is the right one in commenting on the next comment.